[unreadable] [unreadable] The National Cancer Institute (NCI) established the Cancer Genetics Network (CGN) in 1998 as a multi-centered national project designed to support collaborative investigations on the genetic basis of cancer susceptibility and to explore mechanisms to integrate this new knowledge into medical practice. A fundamental issue in this high-risk population is how to optimally screen for early detection of disease. Recent discovery of new biomarkers that change with early stages of disease offer promising non-invasive strategies for screening. However, two aspects that are fundamental for determining an optimal strategy are: for what cancers is the individual at elevated risk and how can longitudinal measurements on a biomarker be used to predict risk of disease. To focus on these issues in the context of the Cancer Genetics Network project, there are two aims in this proposal. The first aim will develop and apply methods for analyzing data on family history to discover what cancers tend to aggregate in individuals and families. We will develop methodology that can be applied to the CGN Registry (family history) data and appropriately accounts for age and mode of ascertainment of participants. We will apply these methods to identify clusters of cancer sites that occur together in families. The second aim of this proposal focuses on the development and application of a method to assess the relationship between longitudinally collected biomarkers and early evidence of disease onset. The method will allow us to analyze data from the CGN Ovarian Cancer (CA125) Biomarker Screening study, as it will appropriately handle this longitudinal and event time data even though the screening schedule for these women varied due to missed or delayed visits. The PI of this proposal is the PI on the Statistical Coordinating Center of the Cancer Genetics Network, and is responsible for overseeing data acquisition and analysis of CGN studies, but is not funded for statistical methods research. This grant would provide the support needed to complete the research for and apply the required statistical methods. This research has the potential of a major impact on the vulnerable population of individuals with family history of cancer, as it will allow the screening strategy to be more focused on cancers to which the individual is at elevated risk, and potentially improve the chances that the disease will be detected in a treatable stage. [unreadable] [unreadable] [unreadable]